Dual Inhibitors of the HIV‐1 IN Catalytic Site and IN−RNA Interactions

While the structural and molecular determinants of ALLINIs efficacy have been extensively characterized, the direct interaction between IN and viral genomic RNA represents the critical link in explaining how these inhibitors impair virion maturation. As previously discussed, this noncatalytic role of IN is indispensable for proper particle morphogenesis. The disruption of the IN–RNA interface by established ALLINIs, such as 11, 17, and 27 (Figures 6 and 17) leads to the formation of noninfectious eccentric particles, which confirms this interface as a strategic therapeutic target [135, 139].

Consistent with this mechanism, recent research has focused on the development of quinolinonyl derivatives as dual‐action inhibitors capable of targeting both the IN catalytic functions and the IN–RNA interaction [148]. This mechanistic study involved: (i) a series of bifunctional DKA quinolinone derivatives and their corresponding ester counterparts, featuring a p‐fluorobenzyl group or a hydrogen atom on the quinolinone nitrogen, and/or alkylamino groups at position 7 of the central core, some of which were previously described as inhibitors of IN catalytic functions [149]; (ii) a set of monofunctional DKA quinolinone derivatives; and (iii) a series of non‐DKA derivatives endowed with various arylmethyloxy groups at 6‐position of the central core, which have been previously described as RNase H inhibitors [150].

Previous findings demonstrated that bifunctional DKA derivatives, exemplified by RDS 1997 [148, 149] (31, Figure 22) effectively inhibited both 3′‐P and ST reactions at nanomolar concentrations, while exhibiting low micromolar activity against HIV‐1 in acutely infected cells [149]. This feature notably differentiates this class of compounds from classical INSTIs, which typically display higher efficacy against ST over 3′‐P. Conversely, compounds endowed with a single DKA branch, including those bearing amino substituents at position 7 of the quinolone core, displayed marked selectivity for ST, showing up to two orders of magnitude higher selectivity over 3′‐P [149, 150, 151]. In addition, non‐DKA derivatives, obtained by shortening the DKA branch into a carboxylic acid function, displayed higher potency against RNase H compared to IN [150].

FIGURE 22: Chemical structures of inhibitors 31–33.

Most of the tested compounds inhibited the catalytic activity of IN in a LEDGF‐independent activity assay (Table 4). In general, DKA derivatives proved more active than non‐DKA analogs, with most of them showing potency within the submicromolar–nanomolar range, and derivative 31 displaying the highest potency (IC50 = 70 nM). SAR studies underscored the critical role of the p‐fluorobenzyl moiety for inhibitory potency, as its removal led to a 50‐fold reduction in potency. In addition, converting the DKA groups into ethyl diketoester moieties reduced activity by nearly 10 times (IC50 = 0.80 µM), indicating the relevance of the acidic portions to activity. In this series, the insertion of amino groups at position 7 of the central core proved a constructive strategy, and indeed 7‐amino substituted diketoester derivatives displayed potency within the submicromolar range. In particular, the insertion of a N,N‐dimethyl group at this position led to an IC50 of 480 nM, while the corresponding bifunctional DKA 32 (Figure 22) was nearly three times more active (IC50 = 181 nM). Interestingly, the monofunctional DKA derivative 33 (Figure 22) bearing a 4‐carboxybenzoyl group in place of the DKA branch at position 6 of the central core showed potency comparable to 31. A similar trend was highlighted also in the ester series, with the ester analog of compound 31 showing similar potency to the ester congener of 33, suggesting that the alternative acidic/ester moiety could mimic the second DKA/diketoester moiety of corresponding bifunctional derivatives.

Compounds | IC 50 LEDGF‐independent IN catalytic activity, μM a | IC 50 IN‐LEDGF binding, μM b | EC 50 for aberrant IN multimerization, μM c | IC 50 IN−RNA binding, μM d | Antiviral efficacy (EC 50 , μM) e | CC 50 , μM f | References
31 | 0.070 ± 0.01 | Qg | Q | 0.84 ± 0.002 | 4.29 (H9) | >200 (U937) | [148, 149]
32 | 0.181 ± 0.06 | AFh | Q | 0.75 ± 0.06 | ND | ND | [148]
33 | 0.0768 ± 0.011 | Q | AF | 1.57 ± 0.11 | ND | ND | [148]

The compounds were also investigated for their ability to inhibit the IN–RNA binding (Table 4). Interestingly, bifunctional derivatives, including both DKA and ester analogs, displayed significant inhibitory potency, with IC50 values within the low‐micromolar‐submicromolar range. The most potent compounds were the bifunctional DKA derivatives 31 (IC50 = 840 nM), along with compound 32 (IC50 = 750 nM) and its ester analog (IC50 = 1.04 µM). Conversely, the removal of the p‐fluorobenzyl moiety from compound 31 led to a 1.5‐fold reduction in potency. A similar trend was highlighted in the ester series, indicating the halobenzyl group as a relevant feature for the inhibition of both IN–RNA binding and IN catalytic activity.

The most promising inhibitors were assessed for their antiviral activity in both the early and late stages of HIV‐1 infection. While derivative 32 and its ester analog only partially inhibited the early steps of HIV‐1 infection without affecting the late steps of viral replication, this result remains consistent with their inhibitory potency toward IN catalytic functions. Differently, a more compelling profile was highlighted by derivative 31 (Figure 23). Indeed, the compound, previously reported to possess antiviral activity against HIV‐1 in acutely infected cells (EC50 = 4.29 μM) [149], displayed antiviral effects on both the early and late stages of HIV‐1infection, with an EC50 of 2.27 µM in the early stage and 40% inhibition at 25 µM in the late stage. Notably, the effects of 31 on both the early and late stages of HIV‐1 infection mirror its multitarget antiviral profile: its potent inhibition of both the 3′‐P and ST processes accounts for its early‐stage effect, while its interference with IN–RNA binding likely contributes to its late‐stage effect.

FIGURE 23: Antiviral activity of compound 31 in the early (A) and late (B) phases of HIV‐infection. Adapted with permission. [148] Copyright 2024, American Chemical Society.

The dual mode of action of 31 was further validated through computational studies [148] (Figure 24). Molecular docking simulations conducted on compound 31 provided a robust structural basis for understanding its multifunctionality as an inhibitor of both the HIV‐1 IN catalytic site and IN–RNA interactions. The study highlighted a clear distinction in structural requirements for catalytic inhibition, revealing that the presence of two DKA branches is essential for potent 3′‐P inhibition (Figure 24A). While one DKA chain chelates a single Mg2+ ion in the orthosteric site, the second branch serves as a critical anchor by forming a salt bridge with K159, which likely traps the ternary complex in a nonfunctional state. In contrast, ST inhibition requires only a single DKA chain to be effective (Figure 24B), as its oxygen atoms can successfully chelate both active‐site Mg2+ cations, explaining why monofunctional derivatives have been reported to only inhibit this specific step [152]. Beyond the catalytic sites, docking simulations suggest that 31 well occupies the CCD–CCD–CTD interface, where it interacts with key basic residues such as K264, R228, and K266 (Figure 24C). Since these residues are critical for RNA binding, this binding mode likely provides a structural explanation for the compound's ability to competitively disrupt interactions between IN and the genomic RNA. Indeed, two of the residues predicted to interact with the p‐fluorobenzyl group and one DKA branch of compound 31 (K264 and K266, respectively) were recently reported to affect the IN–RNA binding through mutagenesis studies [135]. The second DKA arm further stabilizes this allosteric binding via charge‐reinforced hydrogen bonding with the protein backbone.

FIGURE 24: Proposed binding modes of 31. Within the catalytic core, 31 inhibits (A) 3′‐P and (B) ST. In orange, the viral DNA. Magnesium ions are shown as gray spheres. (C) Within the CCD–CTD interface, 31 inhibits the IN−RNA binding.

Besides corroborating the mode of action of 31, this study provides a valuable model for the search of IN inhibitors acting through an innovative mechanism of action. Moreover, based on structural insights from docking studies, chemical modifications to the compound's scaffold can be applied to allow the search for new derivatives endowed with enhanced potency, thereby paving the way for the development of structurally related analogs featuring this innovative mode of IN inhibition.